who can help me? Spiral inductor simulation by HF$$9

need to use lumped ports with grounding layer and define mesh size compared to size of the gap btw turns.

could you give me more specific description about this problem?how to define grounding layer?i'm new in HF$$ world.

gnd layer defined as perfect e-walll at the bottom FASE of silicon die and INFINITE GND is checked.
as well open space is added surround die , because i exactly don't know how to define open space surrounding by default.
also i had change ports rectangles - pay atention to this.

divan,
i wonder why the widths of two lumped ports are not equal?
are there any tricks?

no tricks , dimensions was sticked to the greed and i decide not change them.
this don't matter , you can change to any size approximately compared with the width of the metal strip especially in case of wave port .
the main trick is free space usage surround die.

you have said that the spiral should be finely meshed?
Q factor of the inductor is so high,i think,about 15
but if i use user defined mesh(less than 10um) in the spiral,it takes too long time.futhermore,i don't know if i can get the result i want.
the gap between turns is 1um

sure . this will take time a lot.

It is unnecessary to define mesh size of a dielectric layer proportional to the seperation of metal traces of a spiral inductor, 'cause that will make the simulator divide the whole layer into equally small tetrahedras, and this will waste you a lot of simulation time provided your computer is still able to solve the field. Make use of the adaptive meshing feature of the simulator, let the simulator itself decide which region should be meshed intensively. That is more efficient way, and the simulatin results will be fine, according to my experience.